ارسل ملاحظاتك

ارسل ملاحظاتك لنا







Automatic Extraction Of Ontological Relations From Arabic Text

المصدر: مجلة جامعة الملك سعود - علوم الحاسب والمعلومات
الناشر: جامعة الملك سعود
المؤلف الرئيسي: Al Zamil, Mohammed G. H. (Author)
مؤلفين آخرين: Al Radaideh, Qasem (Co-Author)
المجلد/العدد: مج26, ع4
محكمة: نعم
الدولة: السعودية
التاريخ الميلادي: 2014
الصفحات: 462 - 472
DOI: 10.33948/0584-026-004-011
ISSN: 1319-1578
رقم MD: 973387
نوع المحتوى: بحوث ومقالات
اللغة: الإنجليزية
قواعد المعلومات: science
مواضيع:
كلمات المؤلف المفتاحية:
Arabic Ontology | Lexical Syntactic Patterns | Automatic Extraction Of Relationships
رابط المحتوى:
صورة الغلاف QR قانون
حفظ في:
المستخلص: Automatic extraction of semantic relationships among Arabic concepts to formulate ontology models is crucial for providing rich semantic metadata. Due to the annual increase of Arabic content on the Internet, the need for specialized tools to analyze and understand Arabic text has emerged. This research proposes a methodology that extracts ontological relationships. The goals of this research are: to extract semantic features of Arabic text, propose syntactic patterns of relationships among concepts, and propose a formal model of extracting ontological relations. The proposed methodology has been designed to analyze Arabic text using lexical semantic patterns of the Arabic language according to a set of features. Next, the features have been abstracted and enriched with formal descriptions for the purpose of generalizing the resulted rules. The rules, then, have formulated a classifier that accepts Arabic text, analyzes it, and then displays related concepts labeled with its designated relationship. Moreover, to resolve the ambiguity of homonyms, a set of machine translation, text mining, and part of speech tagging algorithms have been reused. We performed extensive experiments to measure the effectiveness of our proposed tools. The results indicate that our proposed methodology is promising for automating the process of extracting ontological relations.

ISSN: 1319-1578

عناصر مشابهة